Studies on the linetransect census method of the woodland bird populations

III. Differences of the census effectivity between observers

Abstract

1. In the linetransect census data taken at Takizawa and Asahikawa Area, there were no significant differences between the specific relative dominance values calculated from the author's data and that of well-trained young observers (Table 1). This result coinsides with that of Fuji Area (Yui 1974). Namely, these results mean that the census effectivity of every species differs to the same degree between two observers, so long as whose abilities of hearing and seeing are not inferior.
2. The census effectivity of each species was calculated with 95% confidence limit for every two observers in two Areas (Table 2). The majority of the specific census effectivities show the same trend of difference between two observers, although there are almost no significant differences. The differences will become more clear and will converge to the same degree between observers, if more times of census were taken.
3. The levels of census ability of many observers were calculated with relative value against the author's ability, using total recorded numbers of birds per one census standardizing various factors (Table 3). The levels of such observers as well-trained and having normal hearing and seeing ability show about 90±10(%) relative value.
4. To get an accurate estimate of bird density, it is desirable to measure the level of census ability of each observer. But if the error about is 10% permitted, there is no need to measure it, when observer has normal ability in bird census work.